In healthcare practices today, there are opportunities to improve the entire billing lifecycle through the use of Artificial Intelligence (AI) driven software that reduces the redundant administrative workflow and improves reimbursement collections. In reality, AI-based automation is only one part of the equation — human intelligence (HI) is needed to make sense of the process and add the necessary empathy that’s required to make solid decisions.

Whether focusing on pre-care functions, like prior authorizations or insurance coverage discovery, or the back-end process of managing denied claims, advanced automation using AI and machine learning has brought remarkable improvements in speed, accuracy, and the practice bottom line. However, there is still a lingering misunderstanding within the practice management sector about the limitations of AI and the need for human intelligence to synergistically create a viable and comprehensive solution.

How AI + HI Equate to Better Patient Access

The value proposition offered by AI is its ability to complete redundant administrative processes, a critical pain point for practice management, allowing staff members to focus on the patient experience. The human factor comes into play when there are outlying concerns or automation limitations beyond the practice itself such as insurance payers requesting additional clinical documentation or a peer-to-peer review for prior authorization.

Prior Authorization – An AI + HI Powerhouse

While automation has found its way into many of the upfront patient access functions like the insurance verification and patient portion due processes, prior authorization is one area that is still seemingly resistant to change. The 2020 CAQH Index – Closing the Gap: The Industry Continues to Improve, But Opportunities for Automation Remain (CAQH, 2021) reported that a full 79% of providers are still not using a fully automated prior authorization process.

This gap in adoption for prior authorizations may be due to the perception that the complexities of utilization management can’t be resolved by electronic means. But interestingly, it’s here that AI + HI are particularly well suited to work collaboratively.

As an example, a patient arrives at an orthopedic practice with severe osteoarthritis in the left knee requiring knee replacement surgery. In a manual system, this would begin the long and arduous process of filling out forms, faxing, and following up by phone. In a system driven by AI technology, these processes would be completed electronically and in virtual real-time with follow-up happening continuously until completion.

But what happens with the inevitable outliers? As with most things in healthcare, there will always be situations that are more complex and require the nuance of human decision-making. For instance, rapidly evolving insurance guidelines or insurance companies without electronic data transfer capabilities will require HI to take over and make reasoned decisions on the path forward.

Other AI + HI Opportunities in Patient Access and Revenue Cycle Management

While prior authorizations present an enormous opportunity to impact the revenue cycle management (RCM) process through advanced technology, AI + HI is proving to be impactful in several other areas of practice business operations as well.

Insurance Discovery as a Patient Access Tool

The idea of a practice being able to comb a network of clearinghouses, insurance payer direct connections, and public databases to gather undisclosed patient coverage information would have seemed impossible just a few years ago. However, today’s insurance discovery systems can use AI, deep data mining, and probabilistic analytics to glean insurance coverage information that oftentimes even the patient doesn’t know about.

Secondary coverage, Medicaid, Medicare, and commercial insurance can all be accessed either at the beginning of patient treatment or as a final check before a charge is declared uncollectible.

Using AI to Optimize Accounts Receivable

With a recent Medical Group Management Association (MGMA) article reporting that 63% of denials are recoverable on appeal (costing on average $118 each) (MGMA, 2018), it’s easy to see why AI-Driven options are important to increase a practice’s bottom line.

By implementing an AI-driven automation solution that uses machine learning algorithms to forecast the next best action for follow-up and resolution and would include:

  • Forecasting the dollars potentially available based on aging, payer, and modality, plus an estimated timeline.
  • Predictive and deterministic criteria that prioritize activities to focus human intelligence efforts where they are most effective.
  • Automated status checks on claims for cause, e.g., DOS, insurance verification and eligibility data, missing prior authorization, CPT mismatch.
  • Auto-creation of appeal letters.
  • eFax capabilities.
  • Operational analytics to determine root causes and corrections.

In Summary

The old adage of “Man vs. Machine” is obsolete in this new world of advanced technology that’s available to all practices. AI and machine learning offer a powerful value proposition based on an unrivaled ability to process data efficiently and effectively – and healthcare is absolutely swimming in data!

Even with all of this technology, there continues to be the need to incorporate human intelligence to achieve the best overall patient outcomes and to bring a level of nuance and empathy to healthcare and the patient experience.

Navaneeth Nair is Vice President of Products at Infinx Healthcare which provides leading-edge AI-assisted end-to-end solutions across the payment lifecycle, including patient access, prior authorization, and revenue optimization. Navaneeth has over 20 years of experience in healthcare, where he has specialized in leading large-scale technology product and solutions developments.